Noetis engines
One runtime, configurable engines.
Each template is a configuration of specialist responsibilities, tools, evidence rules and an output contract on the shared Noetis runtime. We do not train a new model per customer. Badges show recorded evidence, not plans.
Catalog engines run their specialists on OpenAI models and their claim review and final brief on Anthropic models. Noetis Lite uses Google DeepMind's Gemini for fast reading. Each engine page lists its own models. The way these models are combined comes from research at Encevra Research Labs, so a run is built to go further than a single prompt to any one of them.
Model providers
36 shown